INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_23.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.23.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

_statistics

-0.27

 generate

-0.26

è°¦

-0.25

ìļĶ

-0.25

cum

-0.25

ongyang

-0.24

 insert

-0.24

ilinx

-0.24

 respective

-0.23

å¤´åıĳ

-0.23

POSITIVE LOGITS

éĢŁ

0.28

ETCH

0.25

WON

0.25

Won

0.24

 pronto

0.23

 cottage

0.23

Won

0.23

IBUT

0.23

±Ð¾ÑĤ

0.23

 amor

0.23

Activations Density 0.002%

No Known Activations

This feature has no known activations.